Center for European Studies
Working Paper Series 128
Big Cuts, Little Time:
Welfare State Retrenchment in Sweden
Department of Political Science
P.O. Box 711, 405 30 Göteborg, Sweden
When harsh cuts were introduced in the Swedish welfare state in an agreement between the
centre-right government and the opposition Social Democrats in 1992, there were astonishingly
few disagreements between the political parties as to which social groups should carry the bur-
dens of the cuts. The conventional wisdom on welfare state retrenchment would lead us to expect
a clash of interests, especially considering the strength of interest groups in Sweden and the dif-
ferent constituencies of the five parties included in the agreement. This paper explains why that
did not happen. It argues that the role that key officials played in shaping the 1992 retrenchment
agreement in Sweden was decisive in averting potential political conflicts. In a crisis, politicians
depend on advice from officials as politicians need complex information, often under pressure of
time. This paper argues that key state officials, through their advice, defined both the character of
the crisis and the range of possible solutions. As the number of options was restricted, key offi-
cials were able to define what cuts were reasonable. Within this framework, politicians looked for
practical solutions and, to a large extent, disregarded conflicts of interest. This paper also sug-
gests that the content of such advice depends on what is called the loyalty of key officials, which
depends on the terms of their employment.
When harsh cuts were introduced in the Swedish welfare state in an agreement
between the center-right government and the opposition Social Democrats – the so-
called “first crisis-package” in September 1992 – there were astonishingly few disagree-
ments among the political parties as to which social groups should carry the burdens of
the cuts. Moreover, in contrast with the usual practice, interest groups were largely ex-
cluded from the process. The conventional wisdom on welfare state retrenchment would
lead us to expect a clash of interests, especially considering the strength of interest
groups in Sweden and the different constituencies of the five parties included in the
agreement (Korpi & Palme 2003; Pierson 1996, 2001; Stephens, Huber & Ray 1999). How
can we explain why they could nevertheless reach this agreement?
The main approaches in the literature on welfare state retrenchment do not pro-
vide much help toward answering the question. Scholars following Paul Pierson’s (1994)
influential assessment have drawn attention to the limited range of welfare cutbacks and
claimed that social policy frameworks, although under pressure, remain secure in basi-
cally all mature welfare states. The lack of cutbacks is explained largely by the unpopu-
larity of welfare retrenchment among voters and the influence of new interest groups
such as pensioners, created by the welfare state itself (Esping-Andersen 1999; Green-
Pedersen & Havland 2002; Lindbom 2001; Pierson 1994, 1996, 2001; Taylor-Gooby 2002).
Others have challenged this analysis often using the power-resource approach asso-
ciated with explanations of welfare state expansion. From this perspective, welfare state
cuts are largely explained by the power distribution among class-based actors and ex-
pressed in partisan politics (Allan & Scruggs 2004; Anderson 2001; Clayton & Pontusson
1998; Korpi & Palme 2003; Timonen 2003).
Diverse as these approaches are, they have one assumption in common, and it is
this assumption that is of interest in this study. Both analyze the political process of wel-
fare retrenchment as largely defined by the influence of powerful interest groups, be
they new groups of welfare clients or old class-based groups. Following either of these
approaches one would expect heavy interest group influence. The political parties
should act according to the interest of the strongest interest group in their respective
constituencies, considering their power-resources and their association with the parties.
As already noted the Swedish case suggests that this does not present the full picture.
However, the assumption that interest groups are the decisive factor influencing the out-
comes of retrenchment initiatives evidently does not hold in the case of Sweden in the
This article argues that the role that key officials – that is, centrally placed civil
servants in the Government Offices or agencies – played in shaping the 1992 retrench-
ment agreement in Sweden was decisive in averting potential political conflicts. This ar-
gument builds on Hugh Heclo’s notion that “politics finds its sources not only in power
but also in uncertainty” (Heclo 1974, 305). In a crisis, politicians depend on advice from
officials as politicians need complex information, often under pressure of time. I argue
that key state officials, through their advice, defined both the character of the crisis and
the range of possible solutions. As the number of options was restricted, key officials
were able to define what cuts were reasonable. Within this framework, politicians
looked for practical solutions and, to a large extent, disregarded conflicts of interest.
Without the advice from the key officials, or with other advice, both the welfare cuts and
the process itself would probably have been different. Therefore, the Swedish case sug-
gests that civil servants may play a more important role in determining the extent and
character of cuts in mature welfare states then has been recognized.
This paper also suggests that the content of such advice depends on what I call
the loyalty of key officials. Roughly stated, officials can be loyal either to the state or to
the government, depending on the terms of their employment. In some countries key
state officials depend on the government for their position; the United Kingdom is one
such example. In other countries they do not; Sweden falls into this category. In coun-
tries where state officials depend upon the government, they tend to think like politi-
cians, calculating political gains and losses, therefore including interest considerations in
their advice. In such countries key official advice would not have the disarming effect on
political conflict observed in Sweden. However, in countries where key officials depend
upon the state, political conflicts of interest are disregarded in their advice, and, instead,
the official’s advice is biased by their position in the state apparatus. This is what hap-
pened in Sweden.
Interests, Time, and Influential Officials
Two conclusions may be drawn from the ongoing debates on welfare retrench-
ment. On the one hand, until now, welfare retrenchment has not brought about any
shifts from one welfare regime to another in the mature welfare states due to welfare re-
trenchment (Esping-Andersen 1999; Huber & Stephens 2001; Pierson 1996, 2001; Taylor-
Gooby 2002). The configuration of welfare regimes seems stable, although some analyses
suggest a diminishing role of the welfare state, as social policies do not always cope with
changes involving new social risks (Hacker 2004; Taylor-Gooby 2004). On the other
hand, conclusions concerning stability of welfare state regimes have led to an underesti-
mation of welfare state cuts actually taking place in Oceania, North America and West-
ern Europe. Recent studies show that major cuts have been made in several countries
during the 1980s and 1990s, among them Sweden (Allan & Scruggs 2004; Clayton & Pon-
tusson 1998; Korpi & Palme 2003). It is the variation in welfare cutbacks that this paper
seeks to explore.
As Jacob S. Hacker has noted, too much attention has been paid to what has not
happened to the welfare state and too little to what actually has happened (Hacker 2005).
To some extent, this is due to the tendency in much of the existing literature to define
welfare state retrenchment so as to exclude changes that have not caused major institu-
tional reform. In such studies welfare state retrenchment is seen to occur only if the wel-
fare state no longer remains in the same welfare state regime after the cuts (Lindbom
2001). I instead define retrenchment as any cutbacks in the welfare state, following the
definition of welfare state retrenchment as suggested in the power resources tradition
(Green-Pedersen 2004). There the welfare state is seen as a provider of social rights (Kor-
pi & Palme 2003). Welfare state retrenchment is consequently defined as changes in
these social rights that make them less attractive or generous (Green-Pedersen 2004).
This more inclusive definition is necessary in order to take real changes into account.
The paper shows how cuts in the Swedish welfare state in the 1990s cannot be
explained without taking into account factors typically neglected in studies of the politi-
cal output of retrenchment processes in mature welfare states. The term output refers to
the intended policy result, and should be distinguished from the actual result (outcome).
In particular, I want to show how taking such factors into account helps explain varia-
tions in how the burden of retrenchment is distributed among social groups.
There are several suggestions in existing theory that could – and to a large extent
also do – explain this variation but, as noted earlier, the Swedish case suggests that these